ORIGINAL PAPER

In-tube solid-phase microextraction with a hybrid monolithic columnfor the preconcentration of ultra-trace metals prior to simultaneousdetermination by ICP-MS

Jia-jun Fei1 & Ling-yu Zhao1& Xiao-hong Wu2

& Xiao-bing Cui3 & Hong Min2& Hong-zhen Lian1

& Yi-jun Chen1

Received: 27 January 2020 /Accepted: 15 May 2020# Springer-Verlag GmbH Austria, part of Springer Nature 2020

AbstractThe preparation of an amino-functionalized hybrid monolithic column (TEOS-co-AEAPTES) via one-pot co-condensation oftetraethoxysilane (TEOS) and N-(β-aminoethyl)-γ-aminopropyltriethoxysilane (AEAPTES) in a capillary is descibed. It wasused as solid-phase microextraction (SPME) matrix followed by inductively coupled plasma-mass spectrometry (ICP-MS) fordetermination of trace metals. Under optimum conditions, the amino-functionalized SPME material can simultaneously retainCu(II), Zn(II), Au(III), and Pb(II) with adsorption capacities of 148, 60, 81, and 64μg m−1, respectively. Subsequently, these fourmetal ions can be quantitatively eluted using 1 mol L−1 HNO3 containing 1% thiourea. The retention mechanism of Cu(II),Zn(II), Au(III), and Pb(II) on the amino-functionalized hybrid monolith was explained as the combination of electrostatic andcoordination interactions. With a 10-fold enrichment factor, the calibration curves were established in the range 0.5–100 μg L−1

with linear correlation coefficients above 0.9943 and the limits of quantitation were 0.05 μg L−1 for four target analytes. Thelimits of detection were 0.006, 0.012, 0.004, and 0.007 μg L−1 for Cu(II), Zn(II), Au(III), and Pb(II), respectively. The protocolwas validated by analyzing Certified ReferenceMaterials including standard sediment, soil, and nickel ore, and the results were ingood agreement with their certified values. The relative standard deviations of the method were in the range 0.22–17.6%. Therecoveries of the four metal ions in spiked samples were in the range 88.0–113.8%. Compared to direct ICP-MS determination,the proposed in-tube SPME procedure can effectively eliminate the interference from complex matrix, especially from those oreswith very high content of main metal to improve the accuracy of analysis. Therefore the method is suitable for the simultaneousdetermination of ultra-trace Cu(II), Zn(II), Au(III), and Pb(II) in environmental and mineral samples.

Keywords Hybridmonolithic column . Amino-functionalization . Solid-phasemicroextraction . Tracemetals . ICP-MS

Introduction

Metal elements such as copper and zinc are essential and in-tegral trace elements for life [1]. Lead is a highly toxic elementwhich can accumulate in the organism because of biologicalmagnification in the natural environment [2]. Although gold isscarce and stable in nature [3], long-term use of gold-containing products may have a potential impact on humanhealth [4]. Therefore, it is of great significance to analyzethese trace metals in environmental samples [5]. Besides,since the existence and level of some characteristic elements,whether macro or trace, reflect important region informationof minerals, determination of the abundances of ultra-tracemetals in mineral samples is of critical meaning for theirsource recognition. Because the attention for trace metal ele-ments has been continuously increasing [6], it is quite essen-tial to establish selective, sensitive, and accurate analytical

Electronic supplementary material The online version of this article(https://doi.org/10.1007/s00604-020-04329-0) contains supplementarymaterial, which is available to authorized users.

* Hong-zhen Lianhzlian@nju.edu.cn

* Yi-jun Chencyj-icp@163.com

1 State Key Laboratory of Analytical Chemistry for Life Science,School of Chemistry & Chemical Engineering and Center ofMaterials Analysis, Nanjing University, 163 Xianlin Avenue,Nanjing 210023, China

2 Technical Center for Industrial Product and Raw Material Inspectionand Testing, Shanghai Customs, Shanghai 200135, China

3 College of Pharmacy, Nanjing University of Chinese Medicine,Nanjing 210023, China

https://doi.org/10.1007/s00604-020-04329-0Microchimica Acta (2020) 187: 356

/Published online: 28 May 2020

methods for the simultaneous quantification of multiplemetals in environmental and mineral samples.

Up to now, there are plenty of techniques for the analysis ofmetal elements [7–10]; ICP-MS is most commonly applied inthe determination of trace metals in environmental and miner-al samples due to its advantages of wide linear range, highsensitivity, and simultaneous detection of multiple elements[10]. However, direct analysis of real samples by ICP-MS stillfaces some problems. The concentration of some target metalsin environmental and mineral samples is very low not to bedirectly determined, and that is easily contaminated. On theother hand, the analysis of these environmental and mineralsamples by ICP-MS suffers from mass spectrum and non-mass spectrum interference resulted from matrix effect [11].Isobaric overlap, polyatomic or compound ions, refractoryoxide ions, and double charge ions all cause mass spectruminterference [12]. For mineral samples, a significant problemto surmount is the severe memory effect that is apparent forhigh concentration of main metal of ore in the instrument,leading to non-linear calibration graphs, long washout times,decreasing sensitivity with time, and signals dependent on thematrix. Even worse, the metallic mineral matrix not onlycauses inaccurate measurement results, but also damages massspectrometer and even causes instrument contamination orflameout [13]. Therefore, efficient sample pretreatmentmethods are generally required before ICP-MS analysis forreal environmental and mineral samples.

Solid-phase extraction (SPE) is one of the frequently usedtechniques for the preconcentration of trace metal ions [14].Recently, as a late model solid-phase microextraction extrac-tion (SPME) material, capillary monolithic columns haveattracted extensive attention because of their less sample con-sumption, uniform structure, controllable morphology, con-vective mass transfer, and excellent adsorption performance[15, 16]. The SPME performed within a capillary tube is alsonamed as in-tube SPME or capillary microextraction (CME)[2]. Monolithic columns include inorganic silica [17], organicpolymer [18], and organic-inorganic hybrid monolithic col-umns [19]. Organic-inorganic hybrid monolithic columns areextensively exploited for separation because of their low-swelling, high mass-exchange, large specific surface area,and high mechanical stability [20]. “One-pot” is a facile prep-aration procedure for hybrid monolithic column which meansfunctional groups are introduced directly into the monolith inthe process of co-condensation of organosilane reagents withtetra-alkoxysilanes. Hybrid monolith prepared via one-potcondensation can make functional groups distributed moreevenly and preparation process simpler. Although the applica-tion of organic-inorganic hybrid monolithic columns hasshown great success in organic and biological analysis fields[21–23], only a few works related to them have been conduct-ed in the determination of trace elements and their speciationin environmental samples [24, 25]. Moreover, there are not

any reports to utilize hybrid monolithic material as SPMEmatrix for enriching trace metal elements from high mainmetal ores prior to ICP-MS determination.

In this present work, organic-inorganic hybrid monolithiccolumn (TEOS-co-AEAPTES) prepared via one-pot processby sol-gel technology in a fused capillary was applied forsimultaneous SPME of Cu(II), Zn(II), Au(III), and Pb(II)followed by ICP-MS determination. The adsorption and elu-tion conditions of these four metal ions were optimized com-prehensively. The interaction mechanism between target ionsand amino-functionalized monolith was explained according-ly. The analytical figures of merits were also concluded tofurther prove the reliability of the method. Finally, the appli-cation potential of TEOS-co-AEAPTES monolith was inves-tigated as SPME absorbent for ICP-MS determination ofCu(II), Zn(II), Au(III), and Pb(II) in environmental waters,sediments, soils, coals, and iron ores.

Materials and methods

Reagents and materials

Tetraethoxysilane (TEOS, 98%) was purchased from AlfaAe s a r (Ti a n j i n , Ch i n a ) . N - (β - am inoe t hy l ) -γ -aminopropyltriethoxysilane (AEAPTES, 98%) and thiourea(99+%) were purchased from Ourchem (Shanghai, China).Cetyltrimethylammonium bromide (CTAB, 98+%) was pur-chased from TCI (Tokyo, Japan). Nitric acid was of guaran-teed reagent grade and purchased from Merck (Zurich,Switzerland). All other chemicals were at least of analyticalreagent grade and used without further purification. Pure wa-ter (18.25 MΩ cm) obtained from a Milli-Q water system(Millipore, Bedford, MA, USA) was used throughout theexperiment.

Stock solutions (1 mg mL−1) of Cu(II), Zn(II), and Pb(II)were prepared by dissolving appropriate amounts ofCu(NO3)2·2H2O, Zn(NO3)2·6H2O, and Pb(NO3)2 (NationalInstitute of Metrology, Beijing, China), respectively, in 2%(v/v) diluted HNO3. Stock solution (1 mg mL−1) of Au(III)was prepared by diluting 1000 mg mL−1 of HAuCl4 solution(National Research Center for Certified Reference Materials,Beijing, China). Lower concentration standard solutions wereprepared daily by appropriate dilutions from their stock solu-tions. Analytical mixture standard solution (each of 1 mg L−1)containing Cu(II), Zn(II), Au(III), and Pb(II) was prepared bymixing and diluting the above four stock solutions.

Certified Reference Materials of GBW07309 sediment(GSD-9) and GBW07402 soil (GSS-2) were purchased fromInstitute of Geophysical and Geochemical Prospecting forCertified Reference Materials (Langfang, China). CertifiedReference Material of GBW(E)070110 nickel ore (Fe,46.99%) was purchased from the National Research Center

356 Page 2 of 10 Microchim Acta (2020) 187: 356

for Certified Reference Materials. Tap water was collected inthe Chemical Building in Xianlin Campus of NanjingUniversity. River water was taken from Yangtze River inNanjing. Both water samples were stored with 2% (v/v)HNO3. Soil was taken from Gulou Campus of NanjingUniversity, and sampling of soil is given in the ElectronicSupportingMaterial (ESM). All mineral samples were provid-ed by the Technical Center for Industrial Product and RawMaterial Inspection and Testing, Shanghai Customs(Shanghai, China). Water samples were filtered through a0.45 μm cellulose acetate membrane before analysis.Sediment, soil, coal, and iron ore samples were digested withnitric acid, hydrogen peroxide, and aqua regia before analysis[26]. The detailed procedure is given in the ESM.

Instrumentation

The determination of four target elements including 63Cu,66Zn, 197Au, and 208Pb was performed on a Perkin-ElmerNexION 350D ICP-MS (Perkin-Elmer SCIEX, Concord,Canada) and the optimum operation conditions are summa-rized in Table S1. The pH values of solutions were controlledby a FiveEasy Plus pH meter (Mettler-Toledo, Shanghai,China). During the extraction process, a syringe infusionpump (LSP04-1A, LongerPump, Hebei, China) was used tointroduce solution into the monolithic column.

Scanning electron microscopy (SEM) images were record-ed with a Hitachi S-3400 II SEM (Hitachi, Tokyo, Japan).Fourier-transform infrared (FT-IR) spectra were collected ona NEXUS 870 FT-IR spectrometer (Nicolet, USA) using KBrpellets. Elemental analysis (EA) of the monolith was carriedout using an Elementar Vario EL II elemental analyzer(Elementar, German) with oxygen as the combustion gas.Pore size distribution was investigated by a mercury intrusionporosimeter (Poremaster GT-60, Quantachrome, USA). Thesurface area was calculated using mercury intrusionporosimetry methods.

Preparation of TEOS-co-AEAPTES hybrid monolithiccolumn

The TEOS-co-AEAPTES monolithic capillary was preparedby the procedures reported previously [27, 28] with minormodifications, as detailed in the ESM.

SPME and analysis procedure

The TEOS-co-AEAPTES monolithic capillary was used asSPMEmatrix during Cu(II), Zn(II), Au(III), and Pb(II) enrich-ment procedure (Scheme 1) prior to ICP-MS detection. ThepH of sample solution was adjusted to 4.5 with HNO3 or NH3·H2O. In the extraction step, 3 mL solution (pH 4.5) wasallowed to pump through the TEOS-co-AEAPTES column

at a flow rate of 150 μL min−1. Subsequently, 300 μL ofmixture of 1 mol L−1 HNO3 with 1% (m/v) thiourea was usedas eluent to syringe the analytes retained on the monolithiccapillary at the same flow rate. Both effluents obtained duringthe loading and elution processes were collected and fouranalytes were measured simultaneously by ICP-MS. Purifiedwater was adjusted to pH 4.5 as the blank sample.

Results and discussion

Choice of materials and metal ions

The extraction efficiency of SPE/SPME depends mainly onthe properties of the adsorption materials. According to theLewis acid-base theory, there is a high affinity between tran-sition metal ions as Lewis acid and ligands containing aminoas Lewis base. Therefore, extraction matrices with amino-containing functional group have high adsorption efficiencytowards metal ions [28, 29]. Inspired by this knowledge, threea m i n o - f u n c t i o n a l m o n o m e r s i n c l u d i n gaminopropyltriethoxysilane (APTES), AEAPTES, and [(2-aminoethylamino)ethylamino]propyltrimethoxysilane(AAAPTES) were considered in this study. It was found thatAEAPTES had the best solubility in the solvent system of theexperiment. The obtained monolith prepared by AEAPTESwas uniform and porous, which was suitable for the extractionof metal ions. The octanol-water partition coefficients (logP)of APTES, AEAPTES, and AAAPTES calculated by ACDChemSketch were − 0.22 ± 0.44, − 0.71 ± 0.56, and 1.07 ±0.55, respectively. Therefore, AEAPTES was the most hydro-philic among the three amino-functional monomers, whichsupported the experimental results. As mentioned, hybridmonolith prepared via one-pot condensation can make func-tional groups distributed more evenly and preparation processsimpler. Consequently, TEOS-co-AEAPTES hybrid mono-lithic column prepared via one-pot method was employed toinvestigate the extraction of trace metal ions.

The existence and level of some characteristic elementsincluding Cu(II), Zn(II), Au(III), and Pb(II) prevalent in en-vironmental and mineral samples can reflect important re-gion information. The extraction feasibility of the preparedTEOS-co-AEAPTES hybridmonolithic column towards thecharacteristic metal ions in environmental and mineral sam-ples was investigated. It was found that Cu(II), Zn(II),Au(III), andPb(II) canbequantitatively adsorbed in the sameweakly acidic condition (pH 4.0–6.0). Therefore, these fourions were selected as the target analytes in this study. Thepreparation of the TEOS-co-AEAPTES monolithic columnand theSPMEprocedureofCu(II),Zn(II),Au(III), andPb(II)under optimal conditions are shown in Scheme S1 in theESM.

Microchim Acta (2020) 187: 356 Page 3 of 10 356

Characterization of hybrid monolithic column

As the SEM photographs at different magnifications showed,the TEOS-co-AEAPTES skeletons formed were homoge-neous and also attached to the inner surface of the fused silicacapillary tightly. Even better, no apparent voids along the in-ner wall of capillary could be observed (Fig. 1a, b). Themono-lith possessed continuous macroporous structure andinterconnecting spheres with uniformly small size of particles

and rough surface (Fig. 1c, d). Therefore, the well-controlledskeleton of the monolith was beneficial to the formation ofhigh surface area, low back-pressure, and high permeability.The diameter size of the pore of the TEOS-co-AEAPTESmonolith was analyzed by mercury intrusion porosimetry tobe 3 μm (Fig. S1), and the average surface area was calculated15 m2 g−1. In addition, in order to examine the mechanicalstability of the monolithic capillary, the pressure variation ofthe column was recorded at different flow rates of methanol

Fig. 1 SEM images of the TEOS-co-AEAPTES hybrid monolithiccolumn (a × 150; b × 350; c ×5.00 k; d × 10.0 k)

Scheme 1 Sample pretreatmentand SPME-ICP-MS analysisprocess

356 Page 4 of 10 Microchim Acta (2020) 187: 356

(Fig. S2). The pressure was linearly increased in the testedflow rate range, which revealed that the column had lowback-pressure and excellent mechanical strength.

FT-IR spectrum of the organic-inorganic hybridmonolith isshown in Fig. 2. The absorption bands around 1058 cm−1

corresponded to the stretching vibrations of Si-O-Si groups.The strong peaks around 1472 and 1638 cm−1 resulted fromthe C-N and N-H stretching vibrations, respectively. The ap-pearances of adsorption bands around 2925 and 3375 cm−1

were ascribed to the vibration of C-H and O-H, respectively.All above indicated that the amino-silane group fromAEAPTES had been incorporated into the monolith.Moreover, thermal analysis including TGA and DSC provedthe incorporation of AEAPTES in the matrix (Fig. S3). TGAshowed that the weight loss process could be divided intothree stages, and the proportion of each stage was approxi-mately 7.5%:7.5%:10%, which was consistent with the de-composition of AEAPTES. There were two sharp peaks inthe DSC curve between 200 and 400 °C, which was relatedto the decomposition of amino groups. In addition, EA datashowed the presence of carbon and nitrogen elements derivedfrom organosilane reagents (Table S2). In conclusion, all ofthe characterization results proved that the amino-functionalized monolithic capillary had been successfullysynthesized.

Optimization of analysis procedure

Optimization of adsorption condition

The pH of aqueous solution plays a significant role in adsorp-tion process [21, 24]. In this work, the effect of the pH of 1 mLaqueous solution on the adsorption percentages of 30 μg L−1

Cu(II), Zn(II), Au(III), and Pb(II) on the TEOS-co-AEAPTESmonolithic column was investigated with pH varying in the

range of 1.0–8.0. It could be observed in Fig. 3 that in therelatively wide pH range of 3.0–8.0, four studied metals werequantitatively adsorbed by TEOS-co-AEAPTES monolithiccolumn. At pH below 3.0, the amino-functionalized monolithhad a positively charged surface due to protonation of aminogroups (NH3

+), leading to electrostatic repulsion between ami-no and metal ions including Cu2+, Zn2+ and Pb2+. On thecontrary, NH3

+ on the monolith had certain electrostatic attrac-tion towards AuCl4

−, making partial gold ions absorb on thecolumn.When the pH increased from 3.0 to 8.0, the functionalgroups on the monolith changed from NH3

+ to NH2, the elec-trostatic attraction interaction gradually disappeared, and in-stead, the coordination interaction between NH2 and Cu2+,Zn2+, and Pb2+ dominated. In addition, there was a strongaffinity between NH2 and Au3+, so NH2 could replace Cl−

and coordinate with Au3+, resulting in the adsorption ofAu3+ on the AEAPTES monolith. Since the metal ions mightbe hydrolyzed at the higher pH than 6.0 [30], solution pH wasadjusted to 4.5 for the further experiments.

Other adsorption conditions including the sample flow rateand solution volume [2, 24] were investigated with the resultsshown in Fig. S4 and S5, respectively. In short, the followingexperimental conditions were found to give optimal results:(a) Sample pH value 4.5; (b)Flow rate 150 μL min−1; (c)Solution volume 3 mL.

Optimization of elution condition

Strong acids were selected as the eluents at the beginning inview of the result of dependence of adsorption rate on pH (Fig.3). The result in Fig. 4a showed that either 1 mol L−1 HNO3 orHCl can elute Cu(II), Zn(II), and Pb(II) successfully, exceptAu(III). Due to the electrostatic attraction interaction between

Fig. 3 Adsorption rates of Cu(II), Zn(II), Au(III), and Pb(II) on TEOS-co-AEAPTES monolithic column under different pH. Concentration ofeach metal ions, 30 μg L−1; Sample volume, 1 mLFig. 2 FT-IR spectrum of TEOS-co-AEAPTES monolith

Microchim Acta (2020) 187: 356 Page 5 of 10 356

AuCl4− and the protonated amino groups at very low pH, it

was difficult to release Au(III) from amino-functionalizedmonolith utilizing acids only. Therefore, thiourea was addedto the eluent according to its good affinity to Au(III) [31]. Ascould be seen, the desorption effects of all four metal ionswere excellent when 1% (m/v) thiourea was added in theeluent. In Fig. 4b, there was no obvious difference in thedesorption effect of Cu(II), Zn(II), Au(III), and Pb(II) whilechanging the acid concentration. All these ions could be quan-titatively eluted when the concentration of thiourea varyingfrom 1% to 5% with HNO3/HCl concentration fixing at1 mol L−1 (Fig. 4c). Consequently, 1 mol L−1 HNO3 with1% thiourea was employed for the elution of all target ionsin the following experiments. The effect of elution volume ondesorption rate of target ions was also optimized. A total of1500 μL eluent (1 mol L−1 HNO3 with 1% thiourea) wascontinuously used to desorb Cu(II), Zn(II), Au(III), andPb(II) from the TEOS-co-AEAPTES monolithic column andthe effluent was collected each 300 μL. The result in Fig. 4ddistinctly indicated that the initial 300 μL eluents were plen-tiful for desorbing all target ions from the column.Comprehensive considering the experimental efficiency andenrichment factor, the eluent being as less as better, the elutionvolume was ultimately set to 300 μL when the sample con-sumption was 3 mL with the enrichment of 10 times for targetions. The desorption rates of Cu(II), Zn(II), Au(III), and Pb(II)

did not change obviously during experiment of elution flowrate varying in the range of 20–300 μL min−1 (Fig. S6). Theresult was similar to the flow rate experiment of adsorption.As a result, the elution flow rate we employed in this researchwork was determined as 150 μL min−1.

Adsorption capacity of TEOS-co-AEAPTES monolithiccolumn

Adsorption capacity is a significant indicator to evaluate theperformance of SPME material. The breakthrough curve ex-periment recommended by Hu et al. [32] was carried out inthis present work. In order to investigate the maximum ad-sorption capacity of TEOS-co-AEAPTES monolithic columnfor four target ions, 15 mL solutions containing each targetmetal ion (1 mg L−1, pH 4.5) were continuously pumpedthrough the 5-cm length monol i th ic co lumn at150 μL min−1, respectively. Then, the concentration of targetion in the effluent was determined by ICP-MS. When theadsorption of an ion by the monolithic column was saturated,the concentration of the ion in the effluent suddenly increasedand approached the initial concentration. The breakthroughcurves were shown in Fig. S7 and the maximum adsorptioncapacities of the monolithic column for Cu(II), Zn(II), Au(III),and Pb(II) were 148, 60, 81, and 64 μg m−1 with the relativestandard deviations (RSDs) of 1.0%, 2.2%, 1.6%, and 1.1%,

Fig. 4 Effect of eluent on thedesorption rates of Cu(II), Zn(II),Au(III), and Pb(II). a Effect ofeluent composition; b Effect ofnitric acid or hydrochloric acidconcentration (0.5, 1, and1.5 mol L−1 HNO3/HCl with 1%thiourea); c Effect of thioureaconcentration (1 mol L−1 HNO3/HCl with 1%, 3%, and 5%thiourea); d Effect of elutiontimes, 300 μL eluent (1 mol L−1

HNO3 with 1% thiourea) eachtime. Concentration of each metalions in sample solution,30μg L−1; Sample volume, 3 mL;Solution pH, 4.5; Eluent volumeof a, b, and c, 1 mL

356 Page 6 of 10 Microchim Acta (2020) 187: 356

respectively, which were much higher than subsequent appli-cation needs.

The reproducibility and reusability of TEOS-co-AEAPTES monolithic column

The reproducibility of TEOS-co-AEAPTES monolithic col-umns was evaluated by investigating the recoveries of30 μg L−1 solution containing four target metal ions underthe optimized conditions with six segments of monolithic cap-illary prepared in the same batch and different batches. Theresults in Table S3 indicated that the intra-batch RSDs of thepreparation reproducibility of monolithic column ranged from1.8% (Cu(II)) to 7.3% (Zn(II)) and the inter-batch RSDsranged from 2.4% (Pb(II)) to 8.1% (Zn(II)). Furthermore,the prepared monolithic column can be reused for six timesat least because the recoveries of Cu(II), Zn(II), Au(III), andPb(II) did not decrease obviously during six successive repeat-ing experiments with the RSDs of 4.0%, 8.1%, 0.47%, and1.8%, respectively (Fig. S8). In addition, the stability of thecolumns stored for different duration (freshly prepared, storedfor 1 week, 2 weeks, and 4 weeks) was also evaluated underthe same conditions, and the RSDs (n = 4) of the recoveryrates ranged from 4.8% (Cu(II)) to 7.7% (Zn(II)). The exper-iments verified excellent repeatability and stability of theamino-functionalized hybrid monolithic capillary.

Interference study

The possible coexisting ions in environmental and mineralsamples may compete with the analytes for the active sitesof the amino-functionalized monolith, especially the ore sam-ples that were rich in a variety of high level main metals. Thus,the interference effects of coexisting ions were investigated as

follows. 3 mL aqueous solution containing Cu(II), Zn(II),Au(III), and Pb(II) each at 5 μg L−1 with a certain concentra-tion of coexisting ion was subjected to the general SPME-ICP-MS procedure. The results in Table 1 indicated that therecoveries of four target ions were in the range of 84.3–108.9% in the presence of 25 mg L−1 of Na+, 25 mg L−1 ofK+, 10 mg L−1 of Fe3+, 1 mg L−1 of Ca2+, Al3+, Mg2+, Ni2+,and Mn2+. In addition, main anions such as SO4

2−, NO3−, and

Cl− with the concentration of 10 mg L−1 and PO43− of

1 mg L−1 did not change the recoveries obviously. It demon-strated that the prepared TEOS-co-AEAPTES monolithic col-umn had a relatively good tolerance to the coexisting interfer-ence and maintained adsorption capacity of target ions at ahigh concentration of interfering ions. According to the maincomponents of the real samples that had been digested al-ready, the presence of major interfering ions and their concen-trations were included in this interference study. Furthermore,the important interference components in the real sampleswere focused. For example, the concentration of Fe3+ in theSPME effluent was 18.3% of the initial concentration, whichindicated that the matrix interference could be effectively re-duced by the TEOS-co-AEAPTES monolithic column.Therefore, the SPME-ICP-MS protocol had an excellentanti-interference ability for the determination of ultra-traceCu(II), Zn(II), Au(III), and Pb(II) in environmental and min-eral samples of complex matrices.

Analytical performance of SPME-ICP-MS method

The analytical performance of the developed TEOS-co-AEAPTES monolith based SPME-ICP-MS method was eval-uated under the optimized conditions. The calibration curvesestablished in the range of 0.5–100 μg L−1 for Cu(II), Zn(II),Au(III), and Pb(II) equations were Y = 10.245X − 0.1289, Y =

Table 1 Interference effects ofdiverse ions (mean ± SD, n = 3) Coexisting ion Concentration

(mg L−1)

Recovery (%)

Cu(II) Zn(II) Au(III) Pb(II)

Na+ 25 102.5 ± 3.3 104.0 ± 0.1 94.1 ± 1.3 95.6 ± 1.8

K+ 25 102.7 ± 0.9 95.3 ± 1.5 89.0 ± 1.8 108.9 ± 0.2

Fe3+ 10 93.3 ± 0.4 89.8 ± 1.1 88.4 ± 4.8 95.6 ± 1.0

Mg2+ 1 101.7 ± 3.4 105.0 ± 1.9 84.3 ± 4.3 103.8 ± 0.4

Mn2+ 1 99.4 ± 6.1 103.1 ± 1.4 86.2 ± 1.3 100.7 ± 0.9

Al3+ 1 106.9 ± 0.1 93.4 ± 3.7 91.2 ± 0.8 104.2 ± 2.3

Ni2+ 1 97.1 ± 6.4 103.4 ± 1.4 91.2 ± 3.5 105.4 ± 1.4

Ca2+ 1 97.3 ± 0.9 99.6 ± 9.2 87.7 ± 0.4 105.9 ± 1.8

NO3− 10 96.7 ± 0.9 97.3 ± 1.5 97.0 ± 1.8 93.9 ± 0.2

Cl− 10 102.7 ± 0.9 95.3 ± 1.5 89.0 ± 1.8 108.9 ± 0.2

SO42− 10 102.5 ± 3.3 104.0 ± 0.1 94.1 ± 1.3 95.6 ± 1.8

PO43− 1 107.2 ± 0.2 95.4 ± 2.4 97.9 ± 6.5 97.5 ± 4.2

Microchim Acta (2020) 187: 356 Page 7 of 10 356

9.3065X − 0.2218, Y = 9.7406X + 0.3544, and Y = 9.8710X +0.2647, respectively with r2 of 0.9972, 0.9992, 0.9958, and0.9943, and the limits of quantitation (LOQs) were0.05 μg L−1 for these ions. The limits of detection (LODs,defined as 3-fold the standard deviation of blank signal inten-sity) of Cu(II), Zn(II), Au(III), and Pb(II), with 10-fold enrich-ment factor, were 0.006, 0.012, 0.004, and 0.007 μg L−1, re-spectively for water samples, and 0.008, 0.016, 0.005, and0.009 μg g−1, respectively for sediment, soil, coal, and oresamples. Therefore, this protocol is quite suitable for ultra-trace metal analysis of environmental and mineral samples.

For comparison, the analytical performance of the methodand other related approaches reported in the literature forCu(II), Zn(II), Au(III), and Pb(II) analysis is listed inTable 2. The proposed method has relatively lower LODs

for all target ions and lower than most methods reported.The method can analyze four target ions at the same time withhigh analysis efficiency, while the work about analysis ofAu(III) based on similar materials has been rarely reported.In addition, the proposedmethod can be applied to more kindsof real samples with complex matrices. Although the monolithdid not have the ability to extract a certain metal ion specifi-cally like ion-imprinted materials, and anti-interference testshould be carefully carried out, the proposedmethod can elim-inate the interference from those ores with extremely highcontent of main metal, which is very meaningful for accurateanalysis of multiple characteristic trace ions in environmentaland mineral samples.

The accuracy of the proposed method was verified by de-termining trace Cu(II), Zn(II), Au(III), and Pb(II) in Certified

Table 2 An overview on recently reported nanomaterial-based methods for preconcentration and determination of metal ions

Materials Methods Tolerance levelofFe3+ (mg L−1)

LODs of analyte (μg L−1) Application Ref.

Cu(II) Zn(II) Au(III) Pb(II)

APTES-silica monolithiccolumn

CME-ICP-MS 5 0.006 – – 0.014 Rice flour, human hairand urine

[2]

AAPTS-silica monolithiccolumn

CME-ICP-MS 10 0.002 0.012 – – Human hair, serumand urine

[33]

Down-sized chelatingresin-packed minicolumn

SPE-ICP-MS – 0.0003 0.0007 – 0.0001 Seawater [34]

Nanometer-sized alumina SPE-ICP-MS 2 0.045 0.078 – 0.027 Lake samples, coal flyash,rice flour, rye grass

[35]

TEOS-co-AEAPTES monolithiccolumn

In-tubeSPME-ICP-MS

10 0.006 0.012 0.004 0.007 Water, sediment, soil,coal,iron ore

Thisstudy

Table 3 Contents of Cu(II), Zn(II), Au(III), and Pb(II) in Certified Reference Materials and real environmental and mineral samples

Samples Content (mg kg−1)

Cu(II) Zn(II) Au(III) Pb(II)

Mean ± SD RSD (%) Mean ± SD RSD (%) Mean ± SD RSD (%) Mean ± SD RSD (%)

GSD-9(Certified)

29.8 ± 1.1(32 ± 2)

3.7 72.7 ± 3.1(78 ± 4)

4.3 ND(NA)

– 24.7 ± 0.7(23 ± 3)

2.8

GSS-2(Certified)

17.2 ± 2.7(16.3 ± 0.9)

15.7 44.1 ± 1.7(42 ± 3)

3.9 ND(NA)

– 21.5 ± 1.1(20 ± 3)

5.1

GBW(E)070110(Certified)

126.5 ± 4.5(120 ± 20)

3.6 194.6 ± 14.0(230 ± 20)

7.2 ND(NA)

– 22.2 ± 1.0(23 ± 2)

4.5

Tap water* 5.09 ± 0.23 4.5 48.99 ± 0.11 0.22 ND – 0.98 ± 0.02 2.0

Yangtze River water* 4.96 ± 0.26 5.2 6.56 ± 1.10 0.17 ND – 1.46 ± 0.16 0.11

Soil (Nanjing) 14.6 ± 0.7 4.8 41.0 ± 1.2 2.9 ND – 22.3 ± 1.3 5.8

Russia coal 5.5 ± 0.2 3.6 25.8 ± 0.5 1.9 ND – 9.1 ± 0.1 1.1

Indonesia coal 19.6 ± 0.8 4.1 12.0 ± 0.6 5.0 ND – 1.7 ± 0.3 17.6

Burma iron ore 814.0 ± 28.0 3.4 196.0 ± 17.0 8.7 ND – 83.0 ± 4.0 4.8

Kazakhstan iron ore 733.0 ± 19.0 2.6 446.0 ± 19.5 4.4 ND – 25.5 ± 1.5 5.9

* Concentration, μg L−1

356 Page 8 of 10 Microchim Acta (2020) 187: 356

Reference Materials of GSD-9 (Sediment), GSS-2 (Soil) andGBW(E)070110 (Nickel ore). The results summarized inTable S4 indicated that the determined values of these targetmetals were in accordance with the certified values, whichfurther proved the reliability of this protocol. However, if thenickel ore was directly determined by ICP-MS after digestionand simple dilution without SPME pretreatment, the results offour target elements greatly differed from the certified values(Table S5). This is likely attributed to the strong interferenceof high concentration of main metal ions in the sample solu-tion. It was interestingly found that most of the interfering ionsin the ore could be eliminated by the SPME procedure withthe prepared amino-functionalized hybrid monolith capillary.For example, 82.4% Fe3+ and 78.6% Ni2+ in the digestionsolution were removed after SPME by detecting them in theeffluent. Therefore, this TEOS-co-AEAPTES monolithic col-umn can not only effectively achieve the preconcentration oftarget ions but also powerfully reduce the interference fromthe main metals in ores.

Furthermore, the proposed method was applied to the anal-ysis of environmental water, soil, coal, and iron ore samplesfrom different sources. The determination results for Cu(II),Zn(II), Au(III), and Pb(II) were also listed in Table 3, indicat-ing good precision of the method with the RSDs of 0.22–17.6%. The recoveries of four metal ions spiked in these en-vironmental and mineral samples were summarized inTable S6. The recoveries of four target ions in all spiked sam-ples were obtained 88.0–113.8%, demonstrating that the pro-posed SPME-ICP-MS protocol could be used in the simulta-neous determination of ultra-trace elements in real environ-mental and mineral samples.

Conclusions

TEOS-co-AEAPTES hybrid capillary monolithic column pre-pared via one-pot was employed for the SPME of Cu(II),Zn(II), Au(III), and Pb(II) in environmental and mineral sam-ples. The adsorption mechanism of four metal ions was ex-plained as the combination of electrostatic and coordinationinteractions based on the transformation of NH2 and NH3

+ atdifferent pH. Notably to Au(III), there exists a competingcoordination between amino group and chloride ion underpH 3.0–5.0. The purposed SPME-ICP-MS method possessedstrong anti-interference ability, simple preparation and reus-ability, low sample consumption, and high efficiency of anal-ysis. Although the proposedmethodwas limited in specificity,it can effectively reduce miscellaneous matrix interference incomplex samples, when simultaneous multi-element analysiswas performed. To our knowledge, this work is the first timeto apply organic-inorganic hybrid monolithic capillary for thedetermination of ultra-trace elements in complex

environmental and mineral samples, especially in the oreswith very high content of main metal.

Funding information This work was supported by the National KeyR&D Program of China (2018YFF0215400), the National NaturalScience Foundation of China (21577057, 91643105, 21874065), theNatural Science Foundation of Jiangsu Province (BK20171335), andthe Scientific Research Projects of the General Administration ofCustoms (2019HK074).

Compliance with ethical standards

Conflict of interest The authors declare that they have no conflictof interest.

References

1. Wong CK, Pak AP (2004) Acute and subchronic toxicity of theheavy metals copper, chromium, nickel, and zinc, individuallyand in mixture, to the freshwater copepod Mesocyclopspehpeiensis. Bull Environ Contam Toxicol 73:190–196

2. Zhang L, Chen BB, Peng HY, He M, Hu B (2011)Aminopropyltriethoxysilane-silica hybrid monolithic capillarymicroextraction combined with inductively coupled plasma massspectrometry for the determination of trace elements in biologicalsamples. J Sep Sci 34:2247–2254

3. Medveď J, BujdošM, Matúš P, Kubová J (2004) Determination oftrace amounts of gold in acid-attacked environmental samples byatomic absorption spectrometry with electrothermal atomization af-ter preconcentration. Anal Bioanal Chem 379:60–65

4. Chen ZN, Chen BB, HeM,WangH, Hu B (2019)A porous organicpolymer with magnetic nanoparticles on a chip array forpreconcentration of platinum(IV), gold(III) and bismuth(III) priorto their on-line quantitation by ICP-MS. Microchim Acta 186:107

5. Yu J, Zhu SW, Lu QF, Zhang ZC, Zhang XM, Wang X, Yang W(2018) High sensitive determination of Pb and Zn in refined copperores samples using liquid cathode glow discharge-atomic emissionspectrometry. Spectrosc Spect Anal 38:3550–3557

6. Yang H, He L, Long Y, Li H, Pan S, Liu H, Hu X (2018)Fluorescent carbon dots synthesized by microwave-assisted pyrol-ysis for chromium(VI) and ascorbic acid sensing and logic gateoperation. Spectrochim Acta A 205:12–20

7. Zhou SY, Song N, Liu SX, Chen DX, Jia Q, Yang YW (2014)Separation and preconcentration of gold and palladium ions witha carboxylated pillar arene derived sorbent prior to their determina-tion by flow injection FAAS. Microchim Acta 181:1551–1556

8. Chen AH, Chen SL, Wu YH, Shao Y, Zhang CL (2014) Direct andsequential determination of six metal elements in cooking wine byHR-CS GFAAS. Adv Mater Res 1033-1034:658–662

9. Tang R, Li TP, Gu XS, Li YJ, Yang Y (2010) Studies on six heavymetal elements dissolution characteristics of Andrographis herb byICP-OES. Spectrosc Spect Anal 30:528–531

10. Heitland P, Köster HD (2004) Fast, simple and reliable routinedetermination of 23 elements in urine by ICP-MS. J Anal AtSpectrom 19:1552–1558

11. Huang CY, Beauchemin D (2003) Direct multielemental analysis ofhuman serum by ICP-MSwith on-line standard addition using flowinjection. J Anal At Spectrom 18:951

12. Burney D, Neal CR (2019) Method for quantifying and removingpolyatomic interferences on a suite of moderately volatile elements(Zn, Se, Rb, Ag, Cd, In, Sb, Tl, Pb, and Bi) during solution-modeICP-MS. J Anal At Spectrom 34:1856–1864

Microchim Acta (2020) 187: 356 Page 9 of 10 356

13. Harrington CF, Merson SA, D’Silva TM (2004) Method to reducethe memory effect of mercury in the analysis of fish tissue usinginductively coupled plasma mass spectrometry. Anal Chim Acta505:247–254

14. Chen ZN, Chen BB, He M, Hu B (2019) Magnetic metal-organicf ramework composi tes for dual -column sol id-phasemicroextraction combined with ICP-MS for speciation of tracelevels of arsenic. Microchim Acta 187:48

15. Xu L, Shi ZG, Feng YQ (2011) Porous monoliths: sorbents forminiaturized extraction in biological analysis. Anal Bioanal Chem399:3345–3357

16. Xi Y, Du YX, Sun XD, Zhao SY, Feng ZJ, Chen C, Ding W (2019)A monolithic capillary modified with a copoplymer prepared fromthe ionic liquid 1-vinyl-3-octylimidazolium bromide and styrenefor electrochromatography of alkylbenzenes, polycyclic aromatichydrocarbons, proteins and amino acids. Microchim Acta 187:67

17. Alzahrani E, Fallatah AM (2019) Fabrication of inorganic monolithcoated with gold nanoparticles for protein purification. Int JElectrochem Sci 14:1293–1309

18. Vyviurska O, Lv YQ, Mann BF, Svec F (2018) Comparison ofcommercial organic polymer-based and silica-based monolithiccolumns using mixtures of analytes differing in size and chemistry.J Sep Sci 41:1558–1566

19. Wang Z, Zhao JC, Lian HZ, Chen HY (2015) Aptamer-based or-ganic-silica hybrid affinity monolith prepared via “thiol-ene” clickreaction for extraction of thrombin. Talanta 138:52–58

20. Wang KY, Chen YZ, Yang HH, Li Y, Nie LH, Yao SZ (2012)Modification of VTMS hybrid monolith via thiol-ene click chem-istry for capillary electrochromatography. Talanta 91:52–59

21. Wang M, Ma HF, Chi Q, Li Q, Li M, Zhang HJ, Li CY, Fang HF(2019) A monolithic copolymer prepared from N-(4-vinyl)-benzyliminodiacetic acid, divinylbenzene and N,N'-methylenebisacrylamide for preconcentration of cadmium(II) and cobalt(II)from biological samples prior to their determination by ICP-MS.Microchim Acta 186:537

22. Liu XL, Suo FY, He M, Chen BB, Hu B (2017) Imidazole func-tionalized organic monoliths for capillarymicroextraction of Co(II),Ni(II) and Cd(II) from urine prior to on-line ICP-MS detection.Microchim Acta 184:927–934

23. Weller A, Carrasco-Correa EJ, Belenguer-Sapina C, de los ReyesMauri-Aucejo A, Amoros P, Herrero-Martinez JM (2017) Organo-silica hybrid capillary monolithic column with mesoporous silicaparticles for separation of small aromatic molecules. MicrochimActa 184:3799–3808

24. Zhao LY, Fei JJ, Lian HZ, Mao L, Cui XB (2019) Simultaneousspeciation analysis of chromium and antimony by novel carboxyl-functionalized hybrid monolithic column solid phasemicroextraction coupled with ICP-MS. J Anal At Spectrom 34:1693–1700

25. Zhao LY, Zhu QY, Mao L, Chen YJ, Lian HZ, Hu X (2019)Preparation of thiol- and amine-bifunctionalized hybrid monolithic

column via “one-pot” and applications in speciation of inorganicarsenic. Talanta 192:339–346

26. Santoro A, Held A, Linsinger TPJ, Perez A, Ricci M (2017)Comparison of total and aqua regia extractability of heavy metalsin sewage sludge: the case study of a certified reference material.Trends Anal Chem 89:34–40

27. Yu SB, Geng J, Zhou P, Wang J, Feng AR, Chen XD, Tong H, HuJM (2008) Application of a new hybrid organic-inorganic mono-lithic column for efficient deoxyribonucleic acid purification. AnalChim Acta 611:173–181

28. Li P, ZhangXQ, Chen YJ, LianHZ, Hu X (2014) A sequential solidphase microextraction system coupled with inductively coupledplasma mass spectrometry for speciation of inorganic arsenic.Anal Methods 6:4205–4211

29. Chen S, Hayakawa S, Shirosaki Y, Fujii E, Kawabata K, Tsuru K,Osaka A (2009) Sol-gel synthesis and microstructure analysis ofam ino -mod i f i e d hyb r i d s i l i c a n anopa r t i c l e s f r omaminopropyltriethoxysilane and tetraethoxysilane. J Am CeramSoc 92(9):2074–2082

30. Wang LY, Li JH, Wang JN, Guo XT, Wang XY, Choo J, Chen LX(2019) Green multi-functional monomer based ion imprinted poly-mers for selective removal of copper ions from aqueous solution. JColloid Interf Sci 541:376–386

31. Calle IDL, Cabaleiro N, Costas M, Pena F, Gil S, Lavilla I,Bendicho C (2011) Ultrasound-assisted extraction of gold and sil-ver from environmental samples using different extractants follow-ed by electrothermal-atomic absorption spectrometry. Microchem J97:93–100

32. Hu WL, Hu B, Jiang ZC (2006) On-line preconcentration and sep-aration of Co, Ni and Cd via capillary microextraction on orderedmesoporous alumina coating and determination by inductively plas-ma mass spectrometry (ICP-MS). Anal Chim Acta 572:55–62

33. Zheng F, Hu B (2007) Preparation of a high pH-resistant AAPTS-silica coating and its application to capillary microextraction (CME)of Cu, Zn, Ni, Hg and Cd from biological samples followed by on-line ICP-MS detection. Anal Chim Acta 605:1–10

34. Rahmi D, Zhu YB, Fujimori E, Umemura T, Haraguchi H (2007)Multielement determination of trace metals in seawater by ICP-MSwith aid of down-sized chelating resin-packed minicolumn forpreconcentration. Talanta 72:600–606

35. Yin J, Jiang ZC, Chang G, Hu B (2005) Simultaneous on-linepreconcentration and determination of trace metals in environmen-tal samples by flow injection combined with inductively coupledplasma mass spectrometry using a nanometer-sized alumina packedmicro-column. Anal Chim Acta 540:333–339

Publisher’s note Springer Nature remains neutral with regard to jurisdic-tional claims in published maps and institutional affiliations.

356 Page 10 of 10 Microchim Acta (2020) 187: 356